Identification of unknown metabolites in bamboo leaf extracts by a non-targeted metabolomics approach using UHPLC-QTOF MS/MS driven by chemometrics tools
Bamboo is a rich natural source of promising phytochemicals including flavonoids and other secondary plant metabolites. These metabolites exhibit a broad range of health-promoting effects including anti-inflammatory, anti-oxidant, anti-viral and anti-aging properties. However, there is still a lack of information on the metabolite inventory present in different bamboo species around the world. Thus, there is a great need for detailed information on the metabolite life-stock of different bamboo species influenced by genus, age and geographical origin.
A targeted LC-MS/MS based analysis of major flavonoid composition did not show any correlation regarding anti-oxidative and anti-inflammation properties. Thus, samples of different bamboo leaf extracts of genera Phyllostachys, Fragesia and Sasa as well as young and old leafs were analyzed in a non-targeted metabolic profiling workflow. Results from our non-targeted LC-MS/MS workflow for structure identification of plant metabolites are presented.
The dried bamboo leaves of 25 different species were grinded and 0.5 g of the powder was extracted by accelerated solvent extraction with water/methanol (50/50, v/v, 20 min, 70°C). The extracts were evaporated to dryness with a centrifugal evaporator and 10 mg of residue was cleaned up with solid phase extraction cartridges (1cc, 30 mg Oasis HLB). The purified extract was separated on a Zorbax SB Phenyl column in gradient mode (25-95 B %, A water, B methanol plus 0.1 formic acid) and analyzed with an Agilent 6540 QTOF system in positive mode. Mass spectra were acquired from 100-1000 m/z and with 0 CID voltage for profiling and for structure identification with 10, 20, 40 CID voltage.
The statistical data evaluation was done with mass profiler professional (MPP) and structure identification was supplied with database like dictionary of natural products and metlin.
The investigated bamboo species can be differentiated by principle component analysis of high resolution MS scan data. In addition, for Phyllostachys edulis young and old leaves significant separating features were identified and characterized by the non-targeted metabolic profiling workflow. Non-targeted assessment of the statistical data (with respect to covariance) directed the selection of features for further MS/MS experiments. This results in a focused MS/MS list containing the most differentiating entities. Based on high resolution MS/MS experiments with high mass accuracy (≤ 3ppm) and Product Ion Scan data we structurally identified the most important compounds based on retention time, mass, isotope distribution and fragmentation pattern.
Bamboo metabolite analysis:
Non-targeted workflow with LC-QTOF for bamboo leaves:
PCA score plot of young vs. old Phyllostachys edulis by MPP with Q-TOF data:
LC-MS chromatogram of significant features in young and old Phyllostachys edulis leaves by XCMS
Identification of 2’-Deoxyadenosine as a significant metabolite differentiating young and old Phyllostachys edulis leaves:
Targeted LC-MS/MS analysis allows for direct feed-back loops to optimize natural product extraction procedures.
Non-targeted metabolic profiling in combination with integrated software-directed feature selection was found to be extremely helpful for rapid structural assessment of differentiating features between young and old leaves of Phyllostachys edulis.
The implementation of this integrated workflow accelerates the tedious process from statistical feature identification to structural characterization of relevant metabolites.
Pharmacological studies with the same bamboo extracts exhibit positive effects on the anti-inflammatory and wound-healing assays for young leaves of Phyllostachys edulis.
With the non-targeted metabolite profiling workflow structure identification of significant metabolites is facilitated and allows an interpretation of pharmacological properties on a molecular level.
High resolution MS and MS/MS QTOF data was key to translate differentiating features into structurally characterized metabolites.
Our study provides a model workflow for a comprehensive phytochemical assessment combing high-resolution LC-MS/MS data with pharmacological testing.
[1] Lu B., Wu, X., Tie X., Zhang Y. ; Food and Chemical Toxicology, 2005, 43, 783. [2] Schlotterbeck, G., Hettich, T., Butterweck, V., Riedl, W.; Chimia 2013, 67, 7.
University of Applied Sciences and Arts Northwestern Switzerland, School of Life Sciences, Institute of Chemistry and Bioanalytics, 4132 Muttenz, Switzerland
